 Hi everyone, myself Kalyankova. I'm working in the Overbendwall Center and I'm heading the Cloudify network open source team. Currently, we are working on various open sources like Edge Gallery, Acronome, On-App, Exivella. And today, me and my colleague Gaurav are going to present to you about the Edge Gallery platform, which is open source MAC platform from Huawei. Today's agenda is something like this. Edge Gallery positioning and its scope, technical architecture about the Edge Gallery and the current community progress of the Edge Gallery and the use case demo. FISI B2B ecosystem heavily depends upon the MEC industry application ecosystem prosperity. And in order to achieve this, what we need is open MEC platform that is built over heterogeneous hardware platforms with support of easily integration and quick rollout of MEC applications. And this will result in rich industry application ecosystem and solving the various vertical industry problems in much effective way. And we are trying to bring Edge Gallery in this scope. Currently, there is no unified platform for the innovation achievements in the industry. And there is no closed-loop businesses being formed. And projects are almost impossible to replicate. That's making it more hard. And most of the things needed for MEC like industrial IoT platforms, software capabilities, APS are spread in bits and pieces. And there is a lack of industry platform to build a unified platform, leveraging all these bits and pieces for the enterprise digital transformation to provide an end-to-end solution for various vertical industries. But the core challenge faced by the MEC ecosystem is not changing zero to one or one to 10. Instead, it is how to make it hundreds then how to evolve into 10,000. So there's a big wide gap in this ecosystem to have an unified industrial platform to solve all these problems. Concerning overall requirement and the need of the industry, from Huawei, we have come up with an Edge Gallery platform, which is an MEC platform positioning as a carrier-led Edge Computing architecture in order to accelerate commercial adoption of this platform and resulting in which D2B ecosystem. The idea is to build an open-source Edge Computing platform, which is most compatible with connection and computing in the telco industry, with value propositions like providing IZ native connection capability, adopting Edge native to make Edge services more trustworthy and manageable, and building diversified open ecosystem with support of heterogeneous hardware platforms and quickly unbound and roll out of the applications. And Edge Gallery technology vision is spread across these three important areas, like I mentioned in the previous slide, 5G native connectivity and Edge native architecture and diversified open ecosystem. And 5G native connectivity is nothing but interconnection with the latest 5G network and having 5G network elements in it and 3GPP or ETSA complaint about 5G capability and efficient network data plane and support of latest latency sensitive network like TSN and built-in value-added network function support on this 5G native. On in case of Edge native platform, we want it to be more lightweight resource management, full-stack security mechanism, highly trusted management and operation and support many new Edge native technology like Edge Mesh, Edge AI, Edge Data, Edge SecOps, so many such technologies that can be adopted in this. And on diversified open Edge ecosystem, we are looking at simplified application development and providing good development chain tools to the end user to quickly develop the application and roll out the application and convenient application verification, providing online simulation sandbox and providing a real 5G network, experimental network and unified authentication and intro working and multi ecosystem sharing, like integrating with the public cloud and interrupt with other commercial platforms. So considering all these visions, the Edge gallery is taken into shape. In order to provide the solution for the vertical industry, having quick development and rollout of the application is the need of the hour. The focus is on high value scenario, especially to quickly integrated already available valuable application and end-to-end integration within the days based on end-to-end network application in real environment, end-to-end tool chain, enabling quick application rollout from the lab to the live environment, especially to the production environment. So this is how if you look at it, like the complete sequence like whether it's integrating third party applications or you are developing by yourself, then having it, testing it and certifying it to working on this platform and releasing it to the app repository. So now we'll go in detail about the technical architecture about the Edge gallery. The architecture of Edge gallery is divided into basically five parts. The first part is MEP, that is MEP is autonomous rich in platform and network capabilities, like QoS, DNS, locations, any other the FISI services and also support heterogeneous hardware architectures like X86, RGPU, etc. And the second part is MEC orchestration or application orchestration or application LCM, web caps, policy driven closed loops. And the third part is MECM app repository, which is unified app package format, standardized API between operator app and MECM. And the fourth part is providing MEC development platform, which will provide you the development tool chains, SDK API and sandbox feature for application development and migration, packaging and testing. And the fifth part is application test platform. And this is for automated application certification, test case management, customization and visualizing the test reports. So all three together form the Edge gallery platform. Finally, Edge gallery also is just a cloud connector to implement the interconnection between the open source Edge platform and the public cloud ecosystem, so that applications on the cloud can be deployed at the Edge easily. Currently, Edge gallery plans to share the applications with the public cloud with as in plugins. In future, the plan is to gradually share the applications with other public clouds also. In order to have rich capability of the Edge gallery platform, it's considering top 10 Edge native technologies to support like Edge Infra, Edge Security, Edge Collaboration, Edge Framework, Edge DevSecOps, Edge Orchestrator, like this. So these are all some of them already developed and integrated into the platform like Edge AI, Edge Orchestration, etc. So many others are planned or in pipeline. So there are three technological mega trends and especially when it comes to the lighthouse factory production, connectivity, intelligence and flexible automation. And these are the main principle drivers of the industrial revolution paradigm shift into the smart production. And the lighthouse manufacturers are sprinting ahead with the digital transformation and analytics and changing the rules of the game. Lighthouse factories are implementing advanced manufacturing and the AI-driven technology at a scale and seeing the significant gains. At broader scale of innovation across the world, the change is needed to ensure more winners share this ball. And as shown in this picture, Edge Gallery can perfectly fit here to provide end-to-end solution. So this lighthouse factory, whether it's for the industrial revolution or the next future industrial fire by revolution. So Edge Gallery can fit perfectly in this to leverage all these AI virtual reality, augmented reality, big data, all these new technology it would leverage to provide the end-to-end solution and the vertical industry can benefit from this. Edge Gallery can also be used in industrial vision solution for various quality inspections in the production like appearance inspection, surface quality inspection. So many quality inspections can be done using this one and Edge Gallery perfectly fits into this one. And my colleague Gaurav will cover more in detail on this using one of the use case like a PCB defect detection. Edge Gallery is hatched early last year. From that it has progressed pretty well in very short time of just one year with the rich platform capabilities, rich developer tool chain, security optimization with active contribution from the various community members. So this picture depicts the journey of various releases of the Edge Gallery. Now let's see what is the community behind this Edge Gallery and what is the purpose of this community in order to bring up this Edge Gallery platform. The Edge Gallery opens as communities growing steadily after its first release which is on August last year and this picture depicts successful journey of last one year and many industry partners from various streams like telco operators, public load operators, app developers are joined behind in this community and are actively contributing for this platform. And these are the various members who are part of the Edge Gallery. There are eight senior members and 19 normal members. And Edge Gallery has already integrated, developed nearly 150 applications out of which B2BR around 65, B2C around 85 applications. The major focus of Edge Gallery is on the five industries currently smart campus, industrial manufacturing, transportation and logistic gaming, safety and other industries. So there are already nearly 50 organizations have joined and this opens up and there are nearly 300 people are actively contributing in this community. And Edge Gallery is already a part of that 25, more than 25 innovation incubation based deployment in various provinces in China. Edge Gallery works extensively with upstream and downstream industrial organization bodies for better alignment and adoption, whether it's with LFH or GSMA, ETSI. So it works with various industries so that it can quickly become a de facto standard. It can adapt all those things, new innovations, new changes are quickly adapted into the platform. These are some of the useful links for anybody who would wish to be part of this beautiful journey of the Edge Gallery open source. You are all most welcome to join this open source. Thank you for your time. Now my colleague Gaurav will take over and give the demo of the Edge Gallery use case, especially the PCB defective one. Over to you, Gaurav. Thanks, Kalyan. For next couple of minutes, we are going to see Edge Gallery inaction. Let me start with a brief introduction of the demo. There are two user scenarios supported by Edge Gallery. The first one is an application integration wherein an application vendor wanted to integrate his existing application with Edge Gallery. As part of application integration workflow, first step is to upload an application image and deployment related artifacts to Edge Gallery. Second step is to do a sandbox test wherein an application vendor can leverage Edge Gallery provided sandbox features to carry out his application deployment and test. The next step is a packaging wherein Edge Gallery facilitates auto generation of application package. Then comes the certification test. For this, Edge Gallery provides a flexible means to manage test suits and test cases. Also an application vendor can select some of these test suits against which he wanted to certify his application. Upon successful certification of application, an application is released to an application store. Now at this point of time, an application has been successfully onboarded to an Edge Gallery store. From an application store, application can either be federated to a third party or an operator app store or it could be picked via MEC manager. Now MEC manager carries out the deployment of this application to the desired edges. Also, it does the life cycle management of say the second user scenario is an application development wherein an application vendor wanted to leverage Edge Gallery provided capabilities to carry out his development. The first step in this workflow is a capability selection. So Edge Gallery provides this set of platform capabilities which could be leveraged during an application development. The next step is an application development itself wherein application vendor carries out the development of his application by using the capabilities which he has selected in a previous step. All the latest steps of application development is same with the application integration. Today, as the time is limited, I'll only be demonstrating an application integration flow with the help of a demo PCB defect detection application. As part of this demo, I'm going to show three portals. The first one is a developer platform portal, which includes all the steps from the image upload to the certification. Then there is an application store portal to see all the onboarded application. And at last, there is a MEC manager portal to deploy and to manage an application. So now let me quickly switch to the portals. So this is a developer platform portal. It includes tool chains which would be helpful during the development. It also includes an ability center which has multiple platform capabilities which would be leveraged during the development. These capabilities are there in multiple categories. It includes platform services, network services, AI related video database, etc. A user can get details about any of these services by referring to the service docs which is available. Also, user can try out these APIs by using the online emulator which comes with each and every service. Now to start with a development or an integration, a new project has to be created. As ours is an integration scenario, I will be starting an application integration project. Few of the application details has to be filled in including the application name, version of an application, a provider, a workload type which could be either a container or a VM, an architecture which could be an x86 or an ARM architecture, industry and the type, icon for an application and a brief description. So now a project has been successfully created. The next step is a deployment test as part of which especially an application image has to be uploaded which would be done in two modes. The mode one is to upload an app image to a public repository. A mode two is to upload an app image to a edge gallery provided repository. To save time, I have already pre-uploaded all the required images to edge gallery so I can move on to the next step which is to configure deployment files. This can again be done in two modes. Mode one is to upload a deployment YAML file directly. Mode two is to provide deployment details and the deployment YAML file will be auto-created by the system. For this demo, I have a deployment YAML file available so I will be uploading that. There are set of checks which are performed over an uploaded file including a format check, image information, service information in order to check the correctness of the file. The next step is a deployment test itself as part of which firstly a user has to select a testing environment which could be a 5G environment and a sandbox environment. For this demo, I will be selecting a sandbox environment and starting the deployment. As part of the deployment firstly a deployment files are created then our test node is being assigned after which an application is being instantiated on this test node and at last system waits to get the deployment status of an application. This process is going to take couple of minutes. So we can see the progress bar here. Now currently the application has been instantiated and the system is waiting to get the deployment status. So the application deployment is successful and the deployment result can be viewed down. So in total there are three ports which has been deployed as part of this application. Now a VNC client is also provided using which an application vendor can log in to this test node and carry out further test related to the application. Before moving on to the next step it is a best practice to release the resources. The next step is an application release for which first of all an application configuration has to be provided including an application description file which is nothing but a .md file. Application rules which includes traffic rules and DNS rule can also be added. If this application is a producer application and wanted to list his producing service as part of an ability center he can add the configuration here. For our demo I will be skipping these two steps and I will be saving the configuration upon successfully saving the configuration and application packages generated and the detail of the application package can be seen here. The next step is an application certification. So there are set of test suits which are available for this demo. I will be certifying the application against a community test suit. The community test suit has test cases in three categories. The first one is a compliance test cases. The second category is a sandbox test cases and the third category is a security test cases. Under sandbox test cases the application will be instantiated, terminated and the dependency check would be done. Under the security test cases all the security tests would be performed and the status can be seen above. On successful execution of this test suit user can view and can download the detail certification report. I will be moving on to the next step which is to publish this application to an application store. Okay, so it has been successfully published to an app store and with this we have seen all the steps related to a developer platform portal and I will be switching on to an app store portal. So this is an application store portal. We can move to an app warehouse and we can see that the app which we have published is available as part of the app warehouse. User can download the application package, can get the details about the application, can also provide a comment and score related to this application. As part of my app section user can manage all the onboarded application and as part of apps share the application could be federated to a third party or an operator application store. Now I'll be switching on to the last portal which is an MEC manager portal. So MEC manager is something which is an orchestrator and a manager of Edge Gallery. We can view all the nodes details all the edge nodes which are available in the topology can get details about these edge nodes including their hardware software capabilities and other details. User can also view all the applications which has been instantiated currently there is none. Now to deploy a PCB demo application the first step here is to synchronize the package with MEC manager then we can distribute this particular package to one of the test nodes. The status could be seen here the package is getting distributed to this host IP and we can see that it has been successfully distributed. Now we can deploy or instantiate this application few details needs to be added like an app name app description and in case if there are any hardware capabilities needed by this application. So now application instantiation is in progress to this particular MEC host. We can see that the app has been instantiated successfully and this is the last step of our demo. Now we can see the demo app itself. So this is a demo application. It is an AI based application for PCB defect detection in PCB production line. There are set of reconfigured factory we can select one of it. We can add a production line for this factory associate a camera with it and can start processing. So the processing of this PCB has been started and we can see that each and every PCB is being scanned for the defects. The output could be seen in this window. The red shows that the PCB is defective green shows that there is no defect in the PCB detail results could be viewed and in case there is a defect in a PCB the location and type of defect would be shown by this application. So this is all what I have for today for Edge Gallery demo. Thank you.